Norm and Matrice Proof

43 Views Asked by At

I'm trying to show that the following statement is true:

If $||\mathbf a - \theta \mathbf b||^2 - ||\mathbf a||^2 \geq 0$ for all $\theta \in [0,1]$, then $\mathbf a^T \mathbf b \le 0$.

Is this how I would go about this?? $$\sqrt{a - \theta b}^2 - \sqrt{a}^2 \geq 0$$

$$a- \theta b - a \geq 0$$

I'm not sure how to get to the right conclusion.

(I'm using the Euclidean Norm)

1

There are 1 best solutions below

3
On BEST ANSWER

$$ || a - \theta b||^2 - || a||^2 = || a||^2 + \theta^ 2 || b||^2 - 2\theta a\cdot b - || a||^2 = \theta^ 2 || b||^2 - 2\theta a\cdot b $$

this function of theta is minimum when $$ \theta = \frac{a\cdot b}{|| b||^2} $$

If $ a\cdot b > 0 $ this value is in $[0,1]$.

$$ 0\le f\left(\frac{a\cdot b}{|| b||^2}\right) = -\frac {(a\cdot b)^2}{|| b||^2} $$

This is impossible.